A dynamic multi-generation capacity planning under uncertainties

This research studies multi-generation capacity planning problems under uncertainties. In high-tech industry, because of the frequent introduction of new production technology, capacity planners need to expand their facility while several technology options are available. Oftentimes, capacity of advanced technology can be used to produce products of lower technology. The partial flexibility of capacity makes multi-generation dynamic capacity optimization problems difficult. In this research, the multi-generation capacity planning problems are modeled by dynamic programming. In each decision time, capacity planners can invest in several types of capacity. The objective is to maximize expected revenue over a finite planning horizon. The dynamic programming model is solved by value iteration algorithm (VIA). In numerical study, we verify the robustness of proposed methods by discrete event simulation. Our finding provides general guidelines for multi-generation capacity planning under uncertainties.